基于局部分形特征的快速图像匹配方法

Fast Image Matching Based on Local Fractal Features

利用分形的方法建立了一种新颖的基于分形特征的景像匹配方法．该方法具有良好的抗几何形变能力和较强的抗噪能力．由于利用了分块、序贯检测及分层搜索技术，计算量大为减少，试验结果表明这是一种良好的图像匹配方法．

raditional image matching methods have neither scale invariability nor rotational invariability. A new image matching method based on the fractal characteristics is proposed. With scale invariability and rotational invariability of the fractal parameters, the method is good for resisting geometrical deformation and noise. The computing efficiency is improved because of the adoption of thesegmentation, sequential test and layered searching techniques. Experimental results show that themethod is worth recommendation.